The capacity to misuse X’s Grok artificial intelligence chatbot is once again under scrutiny, with new insights showing that the FBI has been investigating the use of Grok in an ongoing harassment case.
According to a Feb. 25 report from 404 Media in collaboration with Court Watch, court documents show that in January, the FBI obtained a search warrant that required X to provide further insight into how a man who has been accused of stalking and harassment used Grok to generate over 200 fake nude images and videos of a woman.
As per the report: “The details of the investigation are contained in an FBI affidavit about the alleged actions of Simon Tuck, who is accused of extensively harassing and threatening the woman’s husband […] The affidavit notes that, in January, the FBI got a search warrant for the man’s conversations with Grok. The FBI says that it received ‘prompts provided to GrokAI that generated approximately 200 pornographic videos of a woman who closely resembled VICTIM’s wife’s physical appearance.’”
The report also said that the accused man used Grok to generate professional complaints about the woman’s husband, which he then submitted to the husband’s employer.
It’s the latest example of potential misuse of xAI’s chatbot tool, with the company also under investigation in various regions over the use of Grok to general non-consensual nude images.
Earlier this year, the Grok nudification trend saw many users prompting the AI bot to generate nude images from various source photos in the app, with Bloomberg reporting that, at one stage, Grok was generating around 6,700 images every hour that could be categorized as “sexually suggestive of nudifying.”
And despite initially pushing back against calls to make changes to the tool in order to limit such behavior, X eventually moved to restrict Grok’s capacity to generate sexually suggestive and non-consensual nude images.
But that may have come too late, with nudification concerns prompting a range of investigations into X and the safeguards that it has in place (or not) to minimize misuse of its tools. According to Reuters, some users have also found that Grok will still generate sexually suggestive images when prompted, which can have major implications beyond faked images of celebrities.
Which could lead to further avenues for litigation, while also exposing X users to significant risk.
X owner Elon Musk has long railed against regulation, and even claimed that the push is more about censorship than protection.
